While the government increases its efforts to boost the housing market (and the Governor of the Bank of England says he is on the alert for a housing bubble) Andrew Oswald reports on research that shows that high levels of home-ownership cost jobs.
Unemployment is a major source of unhappiness, mental ill-health, and lost income. Yet after a century of economic research on the topic, the determinants of the rate of unemployment are still imperfectly understood, and jobless levels in the industrialized nations are today 10 per cent, with some nations over 20 per cent. The historical focus of the research literature has been on which labour-market characteristics — trade unionism, unemployment benefits, job protection, etc — are particularly influential. With my colleague David G Blanchflower, I propose a different way to think. Our results are relevant to, and a potentially worrying for, a wide range of policy-makers and researchers. The technical version of our work is described in Blanchflower and Oswald (2013).
We argue that the housing market plays a fundamental role as a determinant of the rate of unemployment. Because it makes a good ‘laboratory’, we study modern and historic data for the states of the United States. Our statistical analysis finds the following:
• First, rises in a US state’s home-ownership rate are associated with later rises in joblessness in that state. The effects are strikingly large. A doubling of home-ownership in a state would be associated in the steady state with more than a doubling of the unemployment rate.
• Second, after controlling for state fixed-effects, we show that areas with higher ownership have lower mobility.
• Third, high home-ownership areas have longer commute-to-work times, which can be expected in those areas to raise costs for employers and employees.
• Fourth, high home-ownership areas have lower rates of business formation. It is conceivable — we are not able to offer proof — that this may be due to zoning or NIMBY effects.
We examine the US states as a panel through time and then estimate unemployment equations. Using micro data on millions of randomly sampled Americans, we also estimate equations for the number of weeks worked, the probability of a person being unemployed, the extent of labor mobility, the length of commuting times, and the number of businesses. In this way, our analysis documents statistical links between high levels of home-ownership in a geographical area and high later levels of joblessness in that area. We find that this result is robust across sub-periods going back to the 1980s. The lags from ownership levels to unemployment levels are long. They can take up to five years to be evident. This suggests that high home-ownership gradually interferes in some profound way with the efficient functioning of the labour market.
The data used in our paper are almost wholly from the United States. However, we think our conclusions have wider implications — for Europe and elsewhere. Taken in conjunction with new work by Laamanen (2013), which was done independently of our own , and reaches similar conclusions for the country of Finland, the findings may go some way to explain why nations like Spain (80 per cent owners, 20+ per cent unemployment) and Switzerland (30 per cent owners, 3 per cent unemployment) can have such different mixtures of home-ownership and joblessness. Chart 1 shows that there is a strong positive correlation across the wealthy countries between their home-ownership rates and the latest unemployment rates. Such a chart is open to the sensible criticism that the scatter might be a fluke or an illusion caused by particular countries characteristics.
However, that objection cannot be raised about Chart 2, which is for the United States. It plots very long changes (approximately half-century changes) in home-ownership rates and unemployment rates across the US states — minus Alaska and Hawaii — and generates a similar result. It plots the fifty-year change in home-ownership rates (1950-2000) against a sixty-year change in unemployment rates (1950-2010).
Crucially, the full analysis does not depend on data from the special period of the 2007 US house-price crash; nor does it rely on the idea that home owners are themselves disproportionately unemployed (there is a considerable literature that suggests such a claim is false, or, at best, weak); nor does it imply that spatial compensating differentials theory is incorrect; nor is it Keynesian in spirit; nor does it rest upon the idea of ‘house-lock’ in a housing downturn.
Given the emphasis that most post-war western governments have put on the promotion of home-ownership (one exception is Switzerland, which taxes home-owners’ imputed rents), and the tremendous exchequer cost in tax breaks to having done so, we view our statistical results as deeply worrying for policy-makers. A possible reason why these patterns have attracted so little notice from both researchers and the public is that the time lags are long. High levels of home-ownership do not destroy jobs this year; they tend to do so, on our estimates, some years later. Unless these long linkages are properly understood by researchers and politicians, the deleterious consequences of high levels of home-ownership cannot be appreciated.
What mechanisms lie behind these findings? It is not yet possible to be certain. Our contribution should be seen as a statistical one — as documenting patterns of potential interest to economists and social scientists, and perhaps especially to labor economists, macroeconomists, economic geographers, and urban economists. Nevertheless, we have made an attempt to look below the underlying link between current home-ownership and later joblessness. In doing so, we have found evidence that high home-ownership in a U.S. state is associated with
(i) lower labor mobility,
(ii) longer commutes, and
(iii) fewer new firms and establishments.
It should be emphasized that this is after we have controlled for a wide range of possible confounding influences. Our results are consistent with the recent conclusions of a European study done independently by Laamanen (2013).
Our research does not claim that home owners are unemployed more than renters (very probably they are not). Nor is it an attempt to build on the idea that home owners are less mobile than renters (though they probably are). Instead, because the statistical estimation can control for whether individuals are themselves renters or owners, the patterns documented in the paper are consistent with the idea that the housing market generates negative externalities upon the labour market. Home ownership is dangerous.
Blanchflower, D.G., and Oswald, A.J. 2013. ‘Does high home-ownership impair the labor market?’ NBER working paper.
Laamanen, J-P. 2013. ‘Home ownership and the labour market: Evidence from rental housing market deregulation’. Tampere Economic Working Paper, No. 89, University of Tampere, Finland. Downloadable at http://tampub.uta.fi/handle/10024/68116